Manuscript received May 9, 2023; revised June 5, 2023; accepted July 19, 2023; published January 29, 2024
Abstract—The advent of online learning platforms, such as Coursera and Udemy, has facilitated global access to diverse educational opportunities. However, integrating recommender algorithms into these online learning systems presents a notable challenge. To address this challenge, this research paper proposes an architecture for online course recommender systems that leverages learners’ profiles and learning behavior. The proposed architecture comprises six core components: 1) User Management Engine, 2) Online Course Content Management Engine, 3) Learning Behavior Observer and Recorder, 4) Profile and Learning Behavior Builder Engine, 5) Course Recommendation and Filtering Engine, and 6) Feedback Engine. These components collectively form the foundation of the online course recommender system. In order to facilitate practical implementation, a prototype of the online course recommender system has been developed. To evaluate the system’s performance, a comparative analysis was conducted, to compare the personalized recommended list with the non-personalized recommended list. The analysis revealed an average precision of 77.60% for the personalized recommended list, while the non-personalized recommended list achieved an average precision of 65.60%. These findings highlight the superiority of the personalized approach in generating more accurate and relevant online course recommendations.
Keywords—course recommendation, e-learning, personalization, recommender mechanism, recommender system
Cite: Winyu Niranatlamphong, Wicha Charoensuk, and Worasit Choochaiwattana3, "An Online Course Recommender System in e-Learning Using Learners’ Profile and Learning Behavior-Based Mechanism," International Journal of Information and Education Technology vol. 14, no. 1, pp. 161-167, 2024.